INDEX

Explanations

prone to certain outcomes

The neuron fires on mentions of study population or demographic group labels (e.g. “preteens,” “children,” “teenagers,” “participants,” “men,” etc.).

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 each

-1.05

それぞれ

-0.93

াম

-0.86

Each

-0.86

 occasionally

-0.84

 merely

-0.83

سع

-0.83

 Each

-0.82

so

-0.79

 prevent

-0.79

POSITIVE LOGITS

 tend

2.72

 tended

2.02

 tends

1.98

tend

1.88

 Tend

1.57

Tend

1.51

 cenderung

1.41

 tienden

1.34

 TEND

1.33

are

1.28

Activations Density 0.088%